[Testing the acceptability of applied reference intervals, by using serum creatinine as an example].

: Unfortunately, the international standard ISO 15189:2003 does not give clear guidelines for the choice, development, and testing the acceptability of reference intervals. The so-called normalization of the frequency distribution of serum creatinine concentrations was carried out in the population of outpatients above 18 years of age. Creatinine concentrations were measured, using 5 analytical systems: Abbott Architect (n=9054), Roche Modular (n = 22,947), Roche Integra 400 (n=2748), Roche Integra 700 (n=8350), and Roche Hitachi (n = 20,196). The obtained 95% confidence intervals were compared with the published reference intervals. It is shown that this approach may be valuable in the assessment of acceptability of the used reference intervals, by taking into account the analytical quality of measurement of creatinine concentrations. Whether there is an association of the analytical quality of measurement of creatinine concentrations and their biological variations with the approaches to developing and using the reference intervals in clinical practice is discussed.